Protein Kinases are critical constituents of signal transduction networks and their malfunction is associated with disease including cancer, inflammation, diabetes, and heart disease. Traditional drug discovery efforts have pursued inhibitors that target the ATP binding site with little consideration for other sites on protein kinase surfaces. Our hypothesis is that we can use structure-based design techniques to identify novel small molecule binding sites (exosites) on the surface of protein kinases. We know that other exosites must exist because screens of natural products and synthetic libraries have identified ATP-noncompetitive inhibitors for many members of the kinase family. The first goal of this research is to provide a new therapeutic approach to target three protein kinases, Protein Kinase A, Protein Kinase B and Aurora A Kinase. This will be achieved by the successful completion of three aims (I) To identify, through computational analysis, novel druggable sites ('exosites') on PKA, PKB and Aurora A;(II) To identify, by virtual ligand screening, small drug-like molecules that bind to exosites on these kinases and inhibit function;(III) To develop these inhibitors into drug-leads through chemical optimization, structural characterization and iteration. In our final aim: (IV) We will apply our computational analysis to the whole kinase family and identify druggable exosites for other members. We will make available our findings through the Protein Kinase Resource (a public web-based kinase resource), so that these novel exosites can be targeted by academic groups working on specific protein kinases. The nature of the project is multi-disciplinary and we believe that the goals are within reach due to our combined expertise in computational, molecular and structural biology.